A power quality monitoring method and system for a power conversion package

Abstract

This invention relates to the field of power quality monitoring technology, specifically to a method and system for monitoring the power quality of power conversion equipment. The method includes the following steps: acquiring the output voltage time series signal of the power conversion equipment; and calculating the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integral method. This invention, based on a processing logic combining nonlinear dynamics and a model observer, achieves a deep integration from macroscopic operating mode determination to microscopic fault type localization. It solves the problem of traditional methods' difficulty in distinguishing between internal parameter evolution and external disturbances, improving the accuracy of power quality monitoring and the specificity of fault diagnosis for complex nonlinear power electronic equipment under different operating conditions, and providing multi-dimensional decision-making basis for equipment maintenance.

Description

Technical Field

[0001] This invention relates to the field of power quality monitoring technology, and in particular to a method and system for power quality monitoring of a power conversion equipment. Background Technology

[0002] Power quality monitoring technology involves the real-time acquisition, analysis, and evaluation of physical characteristics of electrical energy in power systems and electrical equipment, such as waveform, frequency, and amplitude, to ensure the reliability and compliance of power supply.

[0003] Existing technologies in practical applications largely rely on linear theories such as Fourier transforms to statistically analyze voltage waveforms, amplitudes, and frequencies. However, when equipment operates in a non-steady-state condition or experiences minute parameter perturbations, traditional linear indicators struggle to capture the chaotic characteristics hidden within the time series and early signs of faults. Due to a lack of modeling and analysis of the internal dynamic mechanisms of the equipment, existing monitoring methods often only focus on whether external indicators at the output port exceed limits. For example, they struggle to distinguish whether voltage distortion is caused by external power grid harmonic interference or by dead-zone effects within the equipment. Furthermore, they fail to detect parameter drift caused by component aging before the voltage amplitude becomes significantly abnormal. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and to propose a power quality monitoring method and system for power conversion equipment.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a power quality monitoring method for a power conversion equipment, comprising the following steps:

[0006] The output voltage time series signal of the power conversion equipment is acquired, and the optimal delay time and optimal embedding dimension of the output voltage time series signal are calculated using the correlation integral method.

[0007] The output voltage time series signal is reconstructed in phase space according to the optimal delay time and the optimal embedding dimension to obtain the reconstructed phase space trajectory.

[0008] The maximum Lyapunov exponent of the reconstructed phase space trajectory is calculated using a small data volume method, and the chaotic operating state of the power conversion equipment is determined based on the maximum Lyapunov exponent.

[0009] A sliding mode observer is constructed for the power conversion equipment, the control input signal of the power conversion equipment is obtained, and the control input signal and the output voltage time series signal are input to the sliding mode observer as the observation objects to obtain the observed estimated voltage.

[0010] Calculate the observation error between the estimated voltage and the time series signal of the output voltage, and generate a perturbation residual signal based on the observation error;

[0011] Component analysis is performed on the disturbance residual signal to extract the DC component features and periodic component features in the disturbance residual signal;

[0012] Based on the DC component characteristics, parameter drift faults of the power conversion equipment are determined; based on the periodic component characteristics, harmonic distortion faults of the power conversion equipment are determined; and a power quality monitoring report is generated in conjunction with the chaotic operating state.

[0013] Preferably, the steps of calculating the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integral method include:

[0014] The output voltage time series signal is divided into several sub-sequences;

[0015] Calculate the correlation integral statistics of the given subsequences;

[0016] Calculate the difference of the correlation integral statistic under different time delays, and determine the time corresponding to the first minimum value of the difference as the optimal delay time;

[0017] Calculate the global minimum value of the correlation integral statistic based on the optimal delay time, and determine the time corresponding to the global minimum value as the optimal time window width;

[0018] The optimal embedding dimension is calculated and determined based on the ratio of the optimal time window width to the optimal delay time.

[0019] Preferably, the step of reconstructing the phase space of the output voltage time series signal based on the optimal delay time and the optimal embedding dimension includes:

[0020] Based on the optimal delay time and the optimal embedding dimension, the one-dimensional output voltage time series signal is mapped to a high-dimensional phase space to generate a phase space state vector.

[0021] The formula for constructing the phase space state vector is:

[0022] ;

[0023] in, Indicates the first A point in phase space, This indicates that the output voltage time series signal is in the first... Voltage values ​​at each sampling time. This represents the number of sampling points corresponding to the optimal delay time. This represents the optimal embedding dimension. Indicates the transpose symbol;

[0024] The reconstructed phase space trajectory is obtained by connecting the phase space state vectors at all times in chronological order.

[0025] Preferably, the steps for calculating the maximum Lyapunov exponent of the reconstructed phase space trajectory using the small data method include:

[0026] Find an initial reference point in the reconstructed phase space trajectory, and calculate the Euclidean distance between the initial reference point and its nearest neighbor.

[0027] The maximum Lyapunov exponent is obtained by tracking the average logarithmic divergence of the Euclidean distance over time and fitting the slope of the straight line of the average logarithmic divergence over time using the least squares method.

[0028] The formula for calculating the average logarithmic divergence is:

[0029] ;

[0030] The maximum Lyapunov exponent is obtained by fitting a linear regression model, and its expression is:

[0031] ;

[0032] in, Indicates the first The average logarithmic distance over discrete time steps This represents the total number of reference points in the reconstructed phase space trajectory. Indicates the first Each reference point and its nearest neighbor point are after passing through Euclidean distance after a time step Represented by natural constant Logarithmic function with base 0. This represents the maximum Lyapunov exponent. Indicates the sampling period. This represents the fitting constant.

[0033] Preferably, the step of determining the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent includes:

[0034] Determine the positive or negative polarity of the maximum Lyapunov exponent;

[0035] If the maximum Lyapunov exponent is negative, the power conversion equipment is determined to be in a stable cycle operation state.

[0036] If the maximum Lyapunov exponent evolves from a negative value to a positive value or remains positive, it is determined that the power conversion equipment is in a state of chaotic precursor operation or chaotic oscillation operation.

[0037] Preferably, the step of constructing the sliding mode observer for the power conversion assembly includes:

[0038] Based on the nominal inductance and nominal capacitance parameters of the power conversion equipment, a nominal state-space equation is established.

[0039] A sign function is introduced as a correction term into the nominal state-space equation to construct the dynamic equation of the sliding mode observer;

[0040] The formula for the dynamic equation is:

[0041] ;

[0042] in, The derivative of the estimated state variable. Represents the system state matrix. This represents the estimated value of the state variable. Represents the input matrix, This refers to the control input signal. Represents the observer gain matrix. Represents a symbolic function. This represents the time-series signal of the output voltage. This represents the output matrix.

[0043] Preferably, the step of calculating the observation error between the observed estimated voltage and the time series signal of the output voltage, and generating a perturbation residual signal based on the observation error, includes:

[0044] Define a sliding surface function, which is the difference between the output voltage time series signal and the estimated value of the state variable at the output port;

[0045] When the sliding mode observer enters the sliding mode, the output value of the sign function is extracted as the equivalent control quantity;

[0046] The equivalent control quantity is filtered using a low-pass filter to obtain the disturbance residual signal containing the total system disturbance information.

[0047] Preferably, the step of performing component analysis on the disturbed residual signal to extract the DC component features and periodic component features in the disturbed residual signal includes:

[0048] Perform a Fast Fourier Transform on the disturbed residual signal to obtain the residual spectrum;

[0049] The amplitude of the component with a frequency of zero in the residual spectrum is extracted as the feature of the DC component;

[0050] The amplitude of the component in the residual spectrum whose frequency is an integer multiple of the fundamental frequency of the output voltage time series signal is extracted as the periodic component feature;

[0051] If the DC component characteristic is greater than the first preset threshold, it is determined that the power conversion equipment has resistance temperature drift or parameter mismatch.

[0052] If the periodic component characteristic is greater than the second preset threshold, it is determined that the power conversion equipment has a dead zone effect or inductor saturation.

[0053] The present invention also provides a system comprising:

[0054] The signal acquisition module is used to acquire the output voltage time series signal of the power conversion equipment, and to calculate the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integration method.

[0055] The phase space reconstruction module is used to reconstruct the phase space of the output voltage time series signal according to the optimal delay time and the optimal embedding dimension to obtain the reconstructed phase space trajectory.

[0056] The chaotic feature analysis module is used to calculate the maximum Lyapunov exponent of the reconstructed phase space trajectory using a small data volume method, and to determine the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent.

[0057] The sliding mode observation module is used to construct a sliding mode observer for the power conversion equipment, acquire the control input signal of the power conversion equipment, and input the control input signal and the output voltage time series signal as the observation objects into the sliding mode observer to obtain the observed estimated voltage.

[0058] The disturbance detection module is used to calculate the observation error between the observed estimated voltage and the output voltage time series signal, and generate a disturbance residual signal based on the observation error;

[0059] The signal component analysis module is used to perform component analysis on the disturbance residual signal and extract the DC component features and periodic component features in the disturbance residual signal.

[0060] The fault diagnosis module is used to determine the parameter drift fault of the power conversion equipment based on the DC component characteristics, determine the harmonic distortion fault of the power conversion equipment based on the periodic component characteristics, and generate a power quality monitoring report in combination with the chaotic operating state.

[0061] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0062] This invention acquires the output voltage time series signal of a power conversion equipment and uses the correlation integral method to calculate the optimal delay time and optimal embedding dimension. The signal is then reconstructed in phase space, mapping the one-dimensional voltage data to a high-dimensional phase space to restore the system's dynamic characteristics. Based on this, the maximum Lyapunov exponent of the reconstructed phase space trajectory is calculated using a small data quantity method. The sign and evolution trend of this exponent are used to accurately determine whether the equipment is in a stable period, a chaotic precursor, or a chaotic oscillation state, thereby capturing the early instability of the nonlinear system. A sliding mode observer is constructed and the actual control input signal of the equipment is introduced as a reference. The state estimation is performed by combining the output voltage to obtain the observed estimated voltage. The observation error between the estimated voltage and the actual output voltage is calculated to generate a disturbance residual signal. This signal is separated into DC component characteristics and periodic component characteristics through low-pass filtering and component analysis. The DC component characteristics are directly related to parameter drift faults such as internal resistance temperature drift or parameter mismatch, while the periodic component characteristics accurately correspond to dead zone effect or harmonic distortion faults caused by inductor saturation. This processing logic based on the combination of nonlinear dynamics and model observer realizes a deep integration from macroscopic operating mode determination to microscopic fault type location. It solves the problem that traditional methods are difficult to distinguish between internal parameter evolution and external disturbances, improves the accuracy of power quality monitoring and the specificity of fault diagnosis for complex nonlinear power electronic equipment under different operating conditions, and provides multi-dimensional decision-making basis for equipment maintenance. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the steps of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0065] Please see Figure 1 This invention provides a technical solution: a method for monitoring the power quality of a power conversion system, comprising the following steps:

[0066] The output voltage time series signal of the power conversion equipment is acquired, and the optimal delay time and optimal embedding dimension of the output voltage time series signal are calculated using the correlation integral method.

[0067] The phase space of the output voltage time series signal is reconstructed based on the optimal delay time and the optimal embedding dimension to obtain the reconstructed phase space trajectory.

[0068] The maximum Lyapunov exponent of the reconstructed phase space trajectory is calculated using the small data quantity method, and the chaotic operating state of the power conversion unit is determined based on the maximum Lyapunov exponent.

[0069] Construct a sliding mode observer for the power conversion equipment, acquire the control input signal of the power conversion equipment, and input the control input signal and the output voltage time series signal as the observation object into the sliding mode observer to obtain the observed estimated voltage.

[0070] Calculate the observation error between the estimated voltage and the output voltage time series signal, and generate a disturbance residual signal based on the observation error;

[0071] Component analysis is performed on the disturbance residual signal to extract the DC component characteristics and periodic component characteristics in the disturbance residual signal;

[0072] Based on the characteristics of the DC component, parameter drift faults of the power conversion equipment are determined, and harmonic distortion faults of the power conversion equipment are determined based on the characteristics of the periodic component. Power quality monitoring reports are generated by combining the chaotic operating state.

[0073] In this embodiment, the steps of calculating the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integral method include: dividing the output voltage time series signal into several subsequences, calculating the correlation integral statistics of several subsequences, calculating the difference of the correlation integral statistics under different time delays, determining the time corresponding to the first minimum value of the difference as the optimal delay time, calculating the global minimum value of the correlation integral statistics based on the optimal delay time, determining the time corresponding to the global minimum value as the optimal time window width, and calculating and determining the optimal embedding dimension based on the ratio of the optimal time window width to the optimal delay time.

[0074] Specifically, the steps for acquiring the output voltage time series signal of the power conversion equipment and calculating the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integration method include: dividing the output voltage time series signal into several sub-sequences and setting the embedding dimension. The search range is 2 to 5, with a set time delay. The search range is 1 to 50, based on the standard deviation of the output voltage time series signal. Set neighborhood radius The range of values, for example, taking for , , and Calculate the correlation integral for each subsequence The correlation integral is represented in the embedding dimension. and neighborhood radius In the lower phase space, the distance is less than The proportion of point pairs is used to construct a statistic using correlation integrals. The statistic reflects the degree of nonlinear correlation of the output voltage time series signal. The difference statistic is obtained by calculating the difference between the maximum and minimum values ​​of the statistic under different neighborhood radii. Calculate the average of the difference statistic across all embedding dimensions. Find the average value With time delay The first minimum point of the variation curve is used to determine the time corresponding to that minimum point as the optimal delay time. Calculate the statistic Average value across all embedding dimensions and neighborhood radii Find the average of the difference statistics with the average of the statistics The global minimum of the sum is used to determine the optimal time window width, and the time corresponding to the global minimum is used as the window width. Using the formula The embedding dimension is calculated, and if the result is not an integer, it is rounded up to obtain the optimal embedding dimension.

[0075] In this embodiment, the step of reconstructing the phase space of the output voltage time series signal based on the optimal delay time and the optimal embedding dimension includes: mapping the one-dimensional output voltage time series signal to a high-dimensional phase space based on the optimal delay time and the optimal embedding dimension, generating a phase space state vector. The formula for constructing the phase space state vector is as follows: ,in, Indicates the first A point in phase space, The output voltage time series signal represents the first... Voltage values ​​at each sampling time. This represents the number of sampling points corresponding to the optimal delay time. Indicates the optimal embedding dimension. The transpose symbol represents the concatenation of the phase space state vectors at all times in chronological order, which yields the reconstructed phase space trajectory.

[0076] Specifically, the steps for reconstructing the phase space of the output voltage time series signal based on the optimal delay time and optimal embedding dimension include: based on the total length of the output voltage time series signal... and optimal delay time and optimal embedding dimension Determine the total number of reconstructed phase space vectors. Data points are extracted sequentially from the output voltage time series signal according to time order. For each starting moment... (in From 1 to ), taking the voltage sample value at that moment as the first component, and every Each sampling point selects the next voltage sample value as the next component, until a full sample is selected. Each component, thus constructing a A column vector of dimension 1 is used as a state point in a high-dimensional phase space. The formula for constructing the phase space state vector is: ,in, Indicates the first A point in phase space, The output voltage time series signal represents the first... Voltage values ​​at each sampling time. This represents the number of sampling points corresponding to the optimal delay time. Indicates the optimal embedding dimension. Indicates the transpose symbol, according to time index. The order of addition will generate The state vectors in the phase space are connected to form a trajectory describing the dynamic evolution of the system, thus obtaining the reconstructed phase space trajectory.

[0077] In this embodiment, the steps for calculating the maximum Lyapunov exponent of the reconstructed phase space trajectory using the small data method include: finding an initial reference point in the reconstructed phase space trajectory, calculating the Euclidean distance between the initial reference point and its neighboring points, tracking the average logarithmic divergence of the Euclidean distance over time, and fitting the slope of the straight line showing the average logarithmic divergence over time using the least squares method to obtain the maximum Lyapunov exponent. The formula for calculating the average logarithmic divergence is as follows: The maximum Lyapunov exponent is obtained by fitting a linear regression model, and its expression is: ,in, Indicates the first The average logarithmic distance over discrete time steps This represents the total number of reference points in the reconstructed phase space trajectory. Indicates the first Each reference point and its nearest neighbor point are after passing through Euclidean distance after a time step Represented by natural constant Logarithmic function with base 0. This represents the maximum Lyapunov index. Indicates the sampling period. This represents the fitting constant.

[0078] Specifically, the steps for calculating the maximum Lyapunov exponent of the reconstructed phase space trajectory using the small data method include: calculating the average period of the output voltage time series signal. In reconstructing the phase space trajectory for each reference point Find its nearest neighbor And apply time separation constraints during the search process. To exclude spurious neighbors with strong time correlation, a reference point is calculated. Nearest Neighbor Euclidean distance at the initial moment Track the evolution of these neighboring points over time. Euclidean distance after each discrete time step For all reference points in phase space at the same time step Logarithmic distance Averaging yields the mean logarithmic divergence sequence. ,exist Follow A linear growth region is selected from the changing curve. The data in this linear region is fitted using the least squares method. The slope of the fitted line is calculated to obtain the maximum Lyapunov exponent. The formula for calculating the mean logarithmic divergence is as follows: The maximum Lyapunov exponent is obtained by fitting a linear regression model, and its expression is: ,in, Indicates the first The average logarithmic distance over discrete time steps This represents the total number of reference points in the reconstructed phase space trajectory. Indicates the first Each reference point and its nearest neighbor point are after passing through Euclidean distance after a time step Represented by natural constant Logarithmic function with base 0. This represents the maximum Lyapunov index. Indicates the sampling period. This represents the fitting constant.

[0079] In this embodiment, the step of determining the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent includes: determining the positive or negative polarity of the maximum Lyapunov exponent; if the maximum Lyapunov exponent is negative, determining that the power conversion equipment is in a stable periodic operating state; if the maximum Lyapunov exponent evolves from a negative value to a positive value or remains positive, determining that the power conversion equipment is in a chaotic precursor operating state or a chaotic oscillation operating state.

[0080] Specifically, the steps for determining the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent include: reading the preset chaotic judgment logic table, comparing the calculated maximum Lyapunov exponent with zero, and if the maximum Lyapunov exponent is less than zero, it indicates that the reconstructed phase space trajectory converges in the phase space over time, and the distance between adjacent trajectories decreases exponentially over time, thus determining that the power conversion equipment is in a stable periodic operating state. If the maximum Lyapunov exponent is greater than zero, it indicates that the reconstructed phase space trajectory is sensitively dependent on the initial conditions, and adjacent trajectories separate exponentially over time, thus determining that the power conversion equipment is in a chaotic precursor operating state or a chaotic oscillation operating state. For the process of gradually changing from negative to positive values, a critical observation window is set. For example, if the exponent shows a monotonically increasing trend for five consecutive calculation cycles and the final value exceeds 0.01, then it is determined that the system is undergoing a phase transition from a periodic state to a chaotic state.

[0081] In this embodiment, the steps for constructing a sliding mode observer for a power conversion system include: establishing a nominal state-space equation based on the nominal inductance and nominal capacitance parameters of the power conversion system; introducing a sign function as a correction term into the nominal state-space equation; and constructing the dynamic equation of the sliding mode observer. The formula for the dynamic equation is as follows: ,in, The derivative of the estimated state variable. Represents the system state matrix. This represents the estimated value of the state variable. Represents the input matrix, Indicates the control input signal. Represents the observer gain matrix. Represents a symbolic function. This represents the output voltage time series signal. This represents the output matrix.

[0082] Specifically, the steps for constructing a sliding mode observer for a power conversion system include: obtaining the circuit topology and component parameters of the power conversion system, including the nominal inductance. nominal capacitance and load resistance Based on Kirchhoff's voltage and current laws, a state equation describing the dynamic characteristics of the system is established. Inductor current and capacitor voltage are selected as state variables, and the system state matrix is ​​constructed. and input matrix For example, for the Buck converter, The elements of the matrix are composed of , and To construct a state observer, a discontinuous switching term is introduced into the observer's feedback loop. This switching term is composed of the sign function of the observation error. The observer gain matrix is ​​then set. This allows the observer's state trajectory to converge to the sliding surface of the system's actual state within a finite time. The dynamic equation is as follows: ,in, The derivative of the estimated state variable. Represents the system state matrix. This represents the estimated value of the state variable. Represents the input matrix, Indicates the control input signal. Represents the observer gain matrix. Represents a symbolic function. This represents the output voltage time series signal. This represents the output matrix.

[0083] In this embodiment, the steps of calculating the observation error between the estimated voltage and the output voltage time series signal, and generating a disturbance residual signal based on the observation error include: defining a sliding surface function, which is the difference between the output voltage time series signal and the estimated state variable value at the output port; when the sliding observer enters the sliding mode, extracting the output value of the sign function as the equivalent control quantity; and using a low-pass filter to filter the equivalent control quantity to obtain a disturbance residual signal containing the total disturbance information of the system.

[0084] Specifically, the steps of calculating the observation error between the estimated voltage and the output voltage time series signal, and generating the perturbation residual signal based on the observation error, include: obtaining the state variable estimates from the sliding mode observer... Extract the component corresponding to the output voltage, calculate the difference between this component and the actual acquired output voltage time series signal, and define the sliding mode surface function. This difference, i.e. When the sliding mode observer runs and the arrival condition is met When the system enters the sliding mode, the sign function is... High-frequency switching occurs to maintain the system's motion on the sliding surface; the output value of the sign function is extracted and multiplied by the observer gain matrix. The high-frequency switching signal containing system disturbance information is obtained. A low-pass filter is designed, and the cutoff frequency of the low-pass filter is set to be greater than the fundamental frequency of the system and less than the switching frequency of the sliding mode observer. For example, the cutoff frequency is set to 10 times the fundamental frequency. The high-frequency switching signal is input into the low-pass filter to filter out the high-frequency jitter component, thereby reconstructing the equivalent control quantity and obtaining the disturbance residual signal containing the total disturbance information of the system.

[0085] In this embodiment, the steps of performing component analysis on the disturbance residual signal and extracting the DC component characteristics and periodic component characteristics in the disturbance residual signal include: performing a fast Fourier transform on the disturbance residual signal to obtain the residual spectrum; extracting the amplitude of the component with a frequency of zero in the residual spectrum as the DC component characteristic; extracting the amplitude of the component with a frequency that is an integer multiple of the fundamental frequency of the output voltage time series signal in the residual spectrum as the periodic component characteristic; if the DC component characteristic is greater than a first preset threshold, it is determined that the power conversion equipment has resistance temperature drift or parameter mismatch; if the periodic component characteristic is greater than a second preset threshold, it is determined that the power conversion equipment has dead zone effect or inductor saturation.

[0086] Specifically, the steps for component analysis of the disturbance residual signal to extract the DC component and periodic component features include: setting the sampling window length to 10 fundamental periods, truncating the disturbance residual signal, performing a Fast Fourier Transform (FFT) using a radix-2 decimation-time algorithm, calculating the amplitude spectrum at each frequency point, locating the spectral line with a frequency of 0Hz in the amplitude spectrum, reading its amplitude as the DC component feature, and locating the frequency equal to the fundamental frequency of the output voltage time series signal in the amplitude spectrum. and its integer multiples ( The system extracts the spectral lines of the power converter and calculates the root mean square (RMS) values ​​of the amplitudes at specific frequency points as periodic component characteristics. A first preset threshold is set, based on the maximum allowable parameter drift of the device (e.g., 1% of the nominal voltage, i.e., 0.24V). A second preset threshold is set, based on the allowable total harmonic distortion (THD) rate of the device (e.g., 2% of the nominal voltage, i.e., 0.48V). The extracted DC component characteristics are compared with the first preset threshold. If the threshold is exceeded, it is determined that there is resistance temperature drift or inductor parameter mismatch. The extracted periodic component characteristics are compared with the second preset threshold. If the threshold is exceeded, it is determined that there is abnormal dead-zone effect of the power device or nonlinear distortion caused by inductor saturation. This confirms that the power converter equipment has a dead-zone effect or inductor saturation.

[0087] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A power quality monitoring method for a power conversion package, characterized by, Includes the following steps: The output voltage time series signal of the power conversion equipment is acquired, and the optimal delay time and optimal embedding dimension of the output voltage time series signal are calculated using the correlation integral method. The output voltage time series signal is reconstructed in phase space according to the optimal delay time and the optimal embedding dimension to obtain the reconstructed phase space trajectory. The maximum Lyapunov exponent of the reconstructed phase space trajectory is calculated using a small data volume method, and the chaotic operating state of the power conversion equipment is determined based on the maximum Lyapunov exponent. A sliding mode observer is constructed for the power conversion equipment, the control input signal of the power conversion equipment is obtained, and the control input signal and the output voltage time series signal are input to the sliding mode observer as the observation objects to obtain the observed estimated voltage. Calculate the observation error between the estimated voltage and the time series signal of the output voltage, and generate a perturbation residual signal based on the observation error; Component analysis is performed on the disturbance residual signal to extract the DC component features and periodic component features in the disturbance residual signal; The parameter drift fault of the power conversion equipment is determined based on the DC component characteristics, the harmonic distortion fault of the power conversion equipment is determined based on the periodic component characteristics, and a power quality monitoring report is generated in combination with the chaotic operating state. The steps for constructing the sliding mode observer for the power conversion system include: Based on the nominal inductance and nominal capacitance parameters of the power conversion equipment, a nominal state-space equation is established. A sign function is introduced as a correction term into the nominal state-space equation to construct the dynamic equation of the sliding mode observer; The formula for the dynamic equation is: ； in, The derivative of the estimated state variable. Represents the system state matrix. This represents the estimated value of the state variable. Represents the input matrix, This refers to the control input signal. Represents the observer gain matrix. Represents a symbolic function. This represents the time-series signal of the output voltage. Indicates the output matrix; The steps of calculating the observation error between the observed estimated voltage and the output voltage time series signal, and generating a perturbation residual signal based on the observation error, include: Define a sliding surface function, which is the difference between the output voltage time series signal and the estimated value of the state variable at the output port; When the sliding mode observer enters the sliding mode, the output value of the sign function is extracted as the equivalent control quantity; The equivalent control quantity is filtered using a low-pass filter to obtain the disturbance residual signal containing the total system disturbance information.

2. The power quality monitoring method of a power conversion package according to claim 1, wherein, The steps for calculating the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integral method include: The output voltage time series signal is divided into several sub-sequences; Calculate the correlation integral statistics of the given subsequences; Calculate the difference of the correlation integral statistic under different time delays, and determine the time corresponding to the first minimum value of the difference as the optimal delay time; Calculate the global minimum value of the correlation integral statistic based on the optimal delay time, and determine the time corresponding to the global minimum value as the optimal time window width; The optimal embedding dimension is calculated and determined based on the ratio of the optimal time window width to the optimal delay time.

3. The power quality monitoring method of a power conversion package according to claim 1, wherein, The step of reconstructing the phase space of the output voltage time series signal based on the optimal delay time and the optimal embedding dimension includes: Based on the optimal delay time and the optimal embedding dimension, the one-dimensional output voltage time series signal is mapped to a high-dimensional phase space to generate a phase space state vector. The formula for constructing the phase space state vector is: ； in, Indicates the first A point in phase space, This indicates that the output voltage time series signal is at the 1st... Voltage value at each sampling time. This represents the number of sampling points corresponding to the optimal delay time. This represents the optimal embedding dimension. Indicates the transpose symbol; The reconstructed phase space trajectory is obtained by connecting the phase space state vectors at all times in chronological order.

4. The power quality monitoring method of a power conversion package according to claim 3, wherein, The steps for calculating the maximum Lyapunov exponent of the reconstructed phase space trajectory using the small data method include: Find an initial reference point in the reconstructed phase space trajectory, and calculate the Euclidean distance between the initial reference point and its nearest neighbor. The maximum Lyapunov exponent is obtained by tracking the average logarithmic divergence of the Euclidean distance over time and fitting the slope of the straight line of the average logarithmic divergence over time using the least squares method. The formula for calculating the average logarithmic divergence is: ； The maximum Lyapunov exponent is obtained by fitting a linear regression model, and its expression is: ； in, Indicates the first The average logarithmic distance over discrete time steps This represents the total number of reference points in the reconstructed phase space trajectory. Indicates the first Each reference point and its nearest neighbor point are after passing through Euclidean distance after a time step Represented by natural constant Logarithmic function with base 0. This represents the maximum Lyapunov exponent. Indicates the sampling period. This represents the fitting constant.

5. The power quality monitoring method of a power conversion package according to claim 4, wherein, The steps for determining the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent include: Determine the positive or negative polarity of the maximum Lyapunov exponent; If the maximum Lyapunov exponent is negative, the power conversion equipment is determined to be in a stable cycle operation state. If the maximum Lyapunov exponent evolves from a negative value to a positive value or remains positive, it is determined that the power conversion equipment is in a state of chaotic precursor operation or chaotic oscillation operation.

6. The power quality monitoring method of a power conversion package of claim 1, wherein, The steps of performing component analysis on the disturbed residual signal and extracting the DC component features and periodic component features from the disturbed residual signal include: Perform a Fast Fourier Transform on the disturbed residual signal to obtain the residual spectrum; The amplitude of the component with a frequency of zero in the residual spectrum is extracted as the feature of the DC component; The amplitude of the component in the residual spectrum whose frequency is an integer multiple of the fundamental frequency of the output voltage time series signal is extracted as the periodic component feature; If the DC component characteristic is greater than the first preset threshold, it is determined that the power conversion equipment has resistance temperature drift or parameter mismatch. If the periodic component characteristic is greater than the second preset threshold, it is determined that the power conversion equipment has a dead zone effect or inductor saturation.

7. The system for power quality monitoring of power conversion equipment according to any one of claims 1-6, characterized in that, include: The signal acquisition module is used to acquire the output voltage time series signal of the power conversion equipment, and to calculate the optimal delay time and optimal embedding dimension of the output voltage time series signal using the correlation integration method. The phase space reconstruction module is used to reconstruct the phase space of the output voltage time series signal according to the optimal delay time and the optimal embedding dimension to obtain the reconstructed phase space trajectory. The chaotic feature analysis module is used to calculate the maximum Lyapunov exponent of the reconstructed phase space trajectory using a small data volume method, and to determine the chaotic operating state of the power conversion equipment based on the maximum Lyapunov exponent. The sliding mode observation module is used to construct a sliding mode observer for the power conversion equipment, acquire the control input signal of the power conversion equipment, and input the control input signal and the output voltage time series signal as the observation objects into the sliding mode observer to obtain the observed estimated voltage. The disturbance detection module is used to calculate the observation error between the observed estimated voltage and the output voltage time series signal, and generate a disturbance residual signal based on the observation error; The signal component analysis module is used to perform component analysis on the disturbance residual signal and extract the DC component features and periodic component features in the disturbance residual signal. The fault diagnosis module is used to determine the parameter drift fault of the power conversion equipment based on the DC component characteristics, determine the harmonic distortion fault of the power conversion equipment based on the periodic component characteristics, and generate a power quality monitoring report in combination with the chaotic operating state. The steps for constructing the sliding mode observer of the power conversion equipment include: Based on the nominal inductance and nominal capacitance parameters of the power conversion equipment, a nominal state-space equation is established. A sign function is introduced as a correction term into the nominal state-space equation to construct the dynamic equation of the sliding mode observer; The formula for the dynamic equation is: ； in, The derivative of the estimated state variable. Represents the system state matrix. This represents the estimated value of the state variable. Represents the input matrix, This refers to the control input signal. Represents the observer gain matrix. Represents a symbolic function. This represents the time-series signal of the output voltage. Indicates the output matrix; The steps of calculating the observation error between the observed estimated voltage and the output voltage time series signal, and generating a perturbation residual signal based on the observation error, include: Define a sliding surface function, which is the difference between the output voltage time series signal and the estimated value of the state variable at the output port; When the sliding mode observer enters the sliding mode, the output value of the sign function is extracted as the equivalent control quantity; The equivalent control quantity is filtered using a low-pass filter to obtain the disturbance residual signal containing the total system disturbance information.

Images